Noisy Chiller: Solutions to Reduce Vibrations

[kublaikhan]

TL;DR Summary

We have water cooled chiller device causing vibrations on a mirror element.
We wish to dampen the vibrations.
Any help/advice will be welcome.

Dear forum members,

We have a water cooled chiller device (see its model in attached link) and we use it to cool a hot mirror element.
Unfortunately the chiller is vibrating and the vibrations couple all the way to the mirror
through a pair of rubber tubes (1cm in diameter, 2m in length) and a copper heat sink.
Needless to say is that we want the mirror to be still as possible.
I tried to firmly attach and "ground" the tubes into a steel post with no profound results.

Any advice on how to dampen the vibrations will be very welcome.

Thanks!
KK

The chiller we have:
http://www.topview-photonics.com/me...tail/data/topview-optical/en/thermotek/t255p/

 

[Asymptotic]

Links at the Topview site appear broken, but I found a copy of the PDF brochure at
https://archive-resources.coleparmer.com/Manual_pdfs/T253P.pdf

and manual at
https://neurophysics.ucsd.edu/Manuals/ThermoTek/T255P_chiller_manual.pdf

The chiller itself is thermoelectric, so whatever vibration you are experiencing is due to the centrifugal water pump.

What type of rubber tubing (Neoprene, natural rubber, etc.)?
Is it 10 mm ID or 10 mm OD?
Are there any kinks in the tubing that impede flow?
Is vibration reduced or eliminated if a loop back hose (manual page 18) is installed in place of the heat sink?

Minimum rated flow is 760 ml/min. Does the copper heat sink limit water flow lower than that?

 

[jrmichler]

Have you discussed this with the chiller manufacturer? That's usually the best place to start. Have you confirmed that the vibrations are from the pump, as opposed to some sort of resonance in the piping? Do you need the full pressure capability of the pump? If not, can the pump be slowed down? Slowing down a centrifugal pump will reduce vibrations caused by the impeller.

Have you measured the frequency of the vibrations, and compared to the impeller vane passing frequency and the RPM of the pump? A vibration at the vane passing frequency can be reduced by a different impeller or by reducing speed. A once per revolution vibration calls for replacing the pump.

Have you looked at getting a different pump? The Grundfos 15-55 circulator pump in my house is so quiet that I have trouble telling if it's running with my ear almost on it. Find your minimum flow rate, and the pump head needed to get that flow rate. Then look at hydronic circulator pumps.

Your profile says PhD student, so you should have access to the people and equipment to do the vibration tests.
If all of that fails, come back and we will talk about air gapping a liquid system.

 

[kublaikhan]

First, thank you for your kind suggestions.

The exact model of the tube is Festo PUN-H-10X1,5, it's made of plastic.
There's one elbow shaped joint connecting tube sections and that's a good point, I will replace it with a straight one.
Unfortunately we don't have the feedback loop anymore. I will try to get one.

Unfortunately the chiller manufacturer didn't respond any of my e-mails.
I guess they care only if you want to buy something new.
I already checked and the pump cannot be slowed down in this model.
This's a good remark, I will definitely need a sensitive accelerometer
to measure the vibrations on the pipe. I'm already trying to get one.
I tried with my phone's accelerometer and it can hardly detect it.
I'm not sure I can replace the pump, besides pumping water this device needs to keep
a steady temperature.

BTW do you have any experience with any kind of dampers (e.g. damping rings),
do you think it will help?

Thanks again!

 

[Dullard]

I'm not sure of the mass of your mirror/heat sink, or how much vibration is too much:
It is entirely possible that the vibration is coupled right through your fluid. I'd start with:

1. Add some mass to your mirror/heat sink. More is better.
2. Add some compressible volume at the pump inlet and outlet, coupled with slight restrictions on the 'mirror' side of those volumes. I'm thinking: Tee in inverted 2 Liter bottles (full of air) and pinch the hoses a bit.

How much either of these 'adjustments' improves things may give some insight on the most profitable way to proceed.

 

[anorlunda]

I'm already trying to get one.
I tried with my phone's accelerometer and it can hardly detect it.

What are your actual requirements? If you need total elimination rather than reduction in vibrations, then no amount of damping will help.

@jrmichler alread menitoned an air gap. That is the most practical way to really eliminate vibrations.

Even scientists need to do engineering in the laboratory. A sound engineering principle is to begin every project by stating the requirements. So, what exactly are your vibration requirements?

 

[Asymptotic]

The exact model of the tube is Festo PUN-H-10X1,5, it's made of plastic.
There's one elbow shaped joint connecting tube sections and that's a good point, I will replace it with a straight one.
Unfortunately we don't have the feedback loop anymore. I will try to get one.

You can make your own feedback loop with a length of tubing and (2) connectors.
How to do it is outlined on page 32 of the manual, but it occurs to me that 10mm may not have been one of the standard chiller connection options (1/4" appears to have been the default).

Were the chiller connectors originally 1/4" then changed to 10mm, or is 1/4" adapted to 10 mm elsewhere?

Colder PLC connector details
https://www.cpcworldwide.com/Product-List/Series/4/Size/8

Does water drain out of the chiller when tubing is disconnected, or not? If it doesn't, then the chiller connectors have internal spring-type check valves similar to this.

I don't have experience with malfunctioning connectors this small, but their larger brethren sometimes become partially jammed and exacerbate flow variations, particularly at low flow rates. Can't say whether it is relevant in this case, but worth keeping in the back of your mind as an unlikely possibility.

Festo PUN-H-10x1.5 is a hard polyurethane (Shore D52; approximately the hardness of shopping cart wheels),
10mm OD , 7mm ID (1.5 mm wall) with a 28 mm minimum bend radius, and 52 mm flow relevant bend radius (that is, flow will be affected if it is less than this). Make sure tubing bends are all greater than this value.

I tried to firmly attach and "ground" the tubes into a steel post with no profound results.

What you might try is to place inline connectors at the steel post, connect the existing tubing to one side of them, and make short tubing runs from them to the mirror heat sink using softer tubing such as silicone rated Shore 50A, or PVC rated Shore 60A hardness. This ought to reduce vibration coupled mechanically through the tubing, although it won't affect fluid pulsations.

 

[Rive]

the chiller is vibrating and the vibrations couple all the way to the mirror through a pair of rubber tubes (1cm in diameter, 2m in length) and a copper heat sink.

It is a possibility that the source of the vibration is a local turbulence in the flow and not any kind of coupling.
Try to modify the flow speed in the heatsink and check if the frequency/amplitude changes or not. Preferably the load on the chiller should stay the same (some kind of 'shortcut' pipe might help).

Also, I would check if the coupling is done through the floor. Maybe a small piezoelectric microphone/disk and an oscilloscope can help in detecting low frequency vibrations.